IRC Archive for channel #xwiki on 22 October 2012

<jvdrean> has left #xwiki
00:19 <vmassol> has joined #xwiki
00:29 <vmassol> has quit
00:56 <vmassol> has joined #xwiki
01:06 <vmassol> has quit
01:33 <vmassol> has joined #xwiki
01:43 <vmassol> has quit
02:10 <vmassol> has joined #xwiki
02:21 <vmassol> has quit
02:46 <vmassol> has joined #xwiki
02:57 <vmassol> has quit
03:13 <cjd> has joined #xwiki
03:24 <vmassol> has joined #xwiki
03:34 <vmassol> has quit
04:00 <vmassol> has joined #xwiki
04:11 <vmassol> has quit
04:25 <sdumitriu> has joined #xwiki
04:37 <vmassol> has joined #xwiki
04:41 <sdumitriu> vmassol: Online?
04:48 <vmassol> has quit
05:12 <tekzilla> has quit
05:14 <tekzilla> has joined #xwiki
05:14 <vmassol> has joined #xwiki
05:21 <sdumitriu> has quit
05:22 <sdumitriu> has joined #xwiki
05:24 <vmassol> has quit
05:51 <vmassol> has joined #xwiki
06:02 <vmassol> has quit
06:13 <sdumitriu> has quit
06:28 <vmassol> has joined #xwiki
06:38 <vmassol> has quit
07:06 <vmassol> has joined #xwiki
07:16 <vmassol> has quit
07:34 <mflorea> has joined #xwiki
07:38 <coud5> has joined #xwiki
07:42 <coud5> has quit
07:43 <vmassol> has joined #xwiki
07:53 <vmassol> has quit
08:14 <vmassol> has joined #xwiki
08:39 <mrvanes> has joined #xwiki
08:40 <sburjan`> has quit
08:41 <sburjan`> has joined #xwiki
08:42 <tmortagne> has joined #xwiki
08:43 <vmassol1> has joined #xwiki
08:44 <vmassol1> good morning
08:44 <vmassol1> release day!
08:44 <vmassol1> let's fix the build ASAP
08:44 <tmortagne> has quit
08:45 <tmortagne> has joined #xwiki
08:45 <vmassol1> fixing clirr excludes in oldcore legacy that I had forgotten
08:47 <vmassol> has quit
09:01 <vmassol1> tmortagne: I think you broke an API
09:01 <vmassol1> (without an exclude)
09:01 <vmassol1> [ERROR] com.xpn.xwiki.objects.ElementInterface: Method 'public boolean apply(com.xpn.xwiki.objects.ElementInterface, boolean)' has been added to an interface
09:01 <vmassol1> is now known as <vmassol>
09:02 <tmortagne> vmassol: I don't understand, I introduced this method some time ago
09:02 <vmassol> yes
09:02 <vmassol> platform has been failing for a long time
09:03 <vmassol> http://ci.xwiki.org/job/xwiki-platform/2750/org.xwiki.platform$xwiki-platform-legacy-oldcore/console
09:03 <vmassol> hmm "only" 19th
09:03 <tmortagne> this method is way older than 19th
09:03 <vmassol> if you introduced it before that then I don't know why it's not been failing before too
09:04 <vmassol> maybe you added an exclude and someone else removed it?
09:04 <tmortagne> maybe someone "cleaned" excludes, looking
09:04 <vmassol> btw I've addd new <excludes> instead of <difference>; I can't find a way to exclude a group of diffs yet, still researching
09:05 <tmortagne> ok
09:05 <vmassol> in your case it should be a <difference>
09:06 <tmortagne> vmassol: I can find <exclude>com/xpn/xwiki/objects/ElementInterface</exclude> in the pom.xml on my side
09:06 <tmortagne> pulling
09:07 <tmortagne> it's still there
09:07 <tmortagne> ha no
09:08 <tmortagne> looks like it has been modified
09:08 <tmortagne> <exclude>com.xpn.xwiki.objects.ElementInterface</exclude>
09:08 <tmortagne> that's why
09:08 <tmortagne> probably a wrong merge
09:10 <tmortagne> caused by rest merge it seems
09:10 <tmortagne> hope nothing else has been broken
09:24 <vmassol> ok
09:24 <vmassol> devs, please focus on the build today
09:25 <vmassol> we need to stabilize ASAP and release
09:26 <sburjan> has joined #xwiki
09:34 <vmassol> @all devs: please don't commit anymore for 4.3M1 (unless blocker bug fix or test fix!) :)
09:43 <Inge-> sorry to detract from release - it looks like backing up xwiki on mySql is rather simple: 1. backup the WEB-INF files, 2. run mysqldump. According to http://platform.xwiki.org/xwiki/bin/view/AdminGuide/Backup.
09:45 <tmortagne> Inge-: there is also permanentDir folder
09:45 <tmortagne> according to http://platform.xwiki.org/xwiki/bin/view/AdminGuide/Backup too ;)
09:45 <tmortagne> unless you don't have any FS attachment and extensions
09:46 <Inge-> right. I thought the attachments were in the DB by default.
09:48 <Inge-> but for extensions etc.. makes a lot of sense to just include it.
09:48 <Inge-> well, that seems easy enough!
09:48 <Inge-> Good luck on the release!
09:49 <tmortagne> Inge-:attachment are in DB by default, that's why I talked about FS attachments
09:51 <Inge-> and then 6 months from now I add extensions, and forget about it.. I'll include it and all will be well.
09:52 <Inge-> then I can just run the debian installer and restore these files again. that should be easy enough to test.
09:52 <tmortagne> yep
09:52 <tmortagne> Inge-: one detail
09:52 <tmortagne> the 4.2 debian insttaller has a bug
09:53 <tmortagne> it does not set the right permanentDir folder in xwiki.properties file
09:53 <tmortagne> (it does not set anything basically so it ends up in tomcat temp ddir)
09:54 <vmassol> has quit
09:54 <Inge-> yes, I changed it to /var/lib/xwiki/data, as you helpfully pointed out last time :
09:54 <Inge-> :)
09:54 <tmortagne> indeed, I forgot
09:54 <Inge-> is lucene indexing on by default, or something I need to set up?
09:54 <Inge-> I see my permanentDir is empty.
09:55 <vmassol> has joined #xwiki
09:55 <tmortagne> it's on by default
09:55 <tmortagne> actually you can't really disable it AFAIK
10:04 <evalica> has joined #xwiki
10:04 <evalica> has quit
10:05 <evalica> has joined #xwiki
10:07 <evalica> has quit
10:09 <Enygma`> has joined #xwiki
10:10 <evalica> has joined #xwiki
10:19 <jvdrean> has joined #xwiki
10:26 <cdop> has joined #xwiki
10:27 <cdop> bon jour
10:28 <vmassol> ok platform is back to ok
10:28 <vmassol> xe should be ok in a few minutes too
10:28 <vmassol> so we can move to fixing the tests
10:28 <vmassol> (functional tests)
10:29 <vmassol> tmortagne, mflorea: could you fix this one which has been failing for a long time: http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-extension/
10:29 <vmassol> (I know it's possibly a flicker but we need to fix it)
10:29 <vmassol> well doesn't look like a flicker fro http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-extension/
10:29 <vmassol> been consistently fialing for the past 10 builds
10:30 <jvdrean> has quit
10:30 <vmassol> should be easy to fix
10:30 <vmassol> expected:<20> but was:<19>
10:30 <vmassol> http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-extension/org.xwiki.enterprise$xwiki-enterprise-test-extension/1142/testReport/org.xwiki.test.ui.extension/ExtensionTest/testPagination/
10:30 <vmassol> maybe a too fragile test?
10:30 <tmortagne> indeed does not looks like a flicker, something probably changed in the environment
10:31 <vmassol> a pity fabio isn't here, his rest commit broke rest build: http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-rest/
10:31 <jvdrean> has joined #xwiki
10:31 <vmassol> fixign it
10:35 <tmortagne> vmassol: who is the release manager btw (I probably missed it) ?
10:36 <vmassol> tmortagne: I'm covering for Fabio, so I'm doing it ATM (Fabio said he could release tomorrow only). So if we can stabilize this morning and before not too late in the afteroon then I'll do the release today
10:36 <tmortagne> ok
10:40 <cdop> where ends up the output from "System.out.println" of a groovy-script in an extension? I got lilith up and running, but it wont catch those messages
10:41 <vmassol> in your console
10:41 <vmassol> actually
10:41 <SvenDowideit> has quit
10:41 <vmassol> in where System.out is configured to go in your servlet container
10:41 <cdop> well, i got no console as tomcat is running the "show"
10:41 <vmassol> with jetty/hsqldb it goes in the console
10:41 <vmassol> did you check catalina.out?
10:41 <cdop> ahh... ok gonna search then where tomcat puts it, thankt
10:42 <cdop> thanks even
10:42 <tmortagne> by default tomact puts it in catalina.out
10:43 <cdop> oh found it :) thx a ton :)
10:44 <SvenDowideit> has joined #xwiki
10:54 <mflorea> has quit
11:03 <vmassol> ok rest tests are passing now :)
11:03 <vmassol> we need to fix http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-selenium/
11:03 <vmassol> and then we're good :)
11:04 <Inge-> do you find selenium/ui tests to be a good thing, or are they brittle and need much management?
11:04 <vmassol> selenium2 (webdriver) is very good
11:04 <vmassol> provided you use a pageobjects strategy and handle ajax well
11:07 <tmortagne> vmassol: I can't reproduce http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-extension/org.xwiki.enterprise$xwiki-enterprise-test-extension/1142/testReport/org.xwiki.test.ui.extension/ExtensionTest/testPagination/ locally
11:07 <tmortagne> not sure what's wrong yet
11:07 <tmortagne> it really does not looks like something that could flicker
11:09 <Inge-> k
11:09 <Slashman> has joined #xwiki
11:15 <mflorea> has joined #xwiki
11:16 <tmortagne> vmassol: it's actuallt back to stable on jenkins...
11:40 <vmassol> tmortagne: I'm fixing another test ATM (AdministrationTest)
11:40 <vmassol> http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-selenium/1864/#showFailuresLink
11:46 <tmortagne> vmassol: fising http://jira.xwiki.org/browse/XWIKI-8276 (again)
11:46 <tmortagne> s/fising/fixing/
11:47 <vmassol> ok, is that the cause of the failing test or something else?
11:57 <tmortagne> vmassol: no
11:58 <tmortagne> nothing to do
11:58 <tmortagne> evalica just reported this bug to me
11:58 <tmortagne> mflorea is trying to fire out how testPagination failing could happen
11:59 <vmassol> ok
12:05 <cdop> Ok, this extension never could have worked: http://extensions.xwiki.org/xwiki/bin/view/Extension/Import+DokuWiki+into+XWiki+Application found the bug in the velocity-code: http://pastebin.com/zK8gjJTy Line 22 to 24... copy-paste-code.
12:05 <cdop> now... how to fix that...
12:25 <tmortagne> has quit
12:41 <mflorea> has quit
12:42 <vmassol> cdop: all I can suggest is to comment on the extension page so that its creator can see it
12:48 <cdop> k, thanks
12:50 <cdop> On "http://www.xwiki.org/xwiki/bin/view/XWiki/ResetPassword" I get "An unknown problem occurred while sending the reset email."
12:51 <vmassol> maybe the smtp server is not set correctly, I'll check
12:51 <cdop> thanks .)
12:54 <vmassol> btw devs, just noticed this in the logs on xwiki.org: https://gist.github.com/3930939
12:54 <vmassol> seems we have some bug
12:56 <vmassol> and another one: https://gist.github.com/3930946
12:57 <vmassol> it's full of errors like this all over the logs
13:18 <sburjan> vmassol: did you do something to my profile ? :)
13:18 <vmassol> nope
13:18 <sburjan> xwiki:XWiki.Sorin was modified by xwiki:XWiki.VincentMassol Generated password reset token - http://www.xwiki.org/xwiki/bin/view/XWiki/Sorin?viewer=changes&amp;rev2=10.5
13:19 <sburjan> don't know what this is. and don't know why you're the author :)
13:24 <vmassol> restarting xwiki.org
13:37 <xwikibot> has joined #xwiki
13:37 <sburjan> vmassol: could you add these blockers also on the RN ? http://jira.xwiki.org/secure/IssueNavigator.jspa?reset=true&jqlQuery=project+%3D+XWIKI+AND+resolution+%3D+Unresolved+AND+priority+%3D+Blocker+ORDER+BY+key+DESC&mode=hide
13:39 <vmassol> I don't rthink 8299 is related to 4.2
13:40 <sburjan> don't know what to say. you can leave it. enough blocker on oracle to stop users
13:40 <vmassol> sburjan: http://www.xwiki.org/xwiki/bin/view/ReleaseNotes/ReleaseNotesXWiki42
13:40 <sburjan> looks cool and sad :)
13:40 <sburjan> but at least it's true and public
13:41 <vmassol> which reminds me that we need to fix the oracle issue for 4.3 and nobody is working on that
13:41 <vmassol> ...
13:42 <vmassol> new test failure: http://ci.xwiki.org/job/xwiki-enterprise-test-escaping/org.xwiki.enterprise$xwiki-enterprise-test-escaping/1878/
13:42 <vmassol> any idea anyone?
13:51 <vmassol> looks good
13:56 <sburjan> yes, but xwiki.org is a little bit slow on my side
13:57 <sburjan> OR I can't edit objects my user account :)
13:57 <vmassol> that's normal
13:57 <vmassol> it doesn't mean xwiki.org is slow
13:57 <vmassol> it's actually quite fast now
13:58 <vmassol> the reason is that you have rights objects in your profile
13:58 <sburjan> I guess it's because of rights
13:58 <sburjan> yes ..
13:58 <vmassol> and editing your profile means
13:58 <vmassol> loading all users in memory
13:58 <vmassol> :(
13:59 <vmassol> we're talking about 4193 users
13:59 <cdop> ok, i read last weak something, that xwiki is only single-threaded (due to pygments iirc), still like that?
13:59 <vmassol> removing some spam users
13:59 <vmassol> cdop: not xwiki
13:59 <vmassol> just the code macro
14:00 <cdop> ah, as long as the page doesnt have the code-macro included, multi-threaded view possible, right?
14:00 <vmassol> cdop: yes
14:00 <cdop> ok, thanks
14:00 <vmassol> we need to raise the issue btw
14:00 <vmassol> doing it now
14:01 <cdop> :)
14:04 <vmassol> done, reported XRENDERING-249
14:10 <mrvanes> has quit
14:12 <sburjan`> has quit
14:17 <mflorea> has joined #xwiki
14:17 <Slashman_> has joined #xwiki
14:23 <sburjan> vmassol: does you JIRA Macro version support inline blocks ?
14:24 <vmassol> yes
14:24 <vmassol> that's not hte problem
14:24 <vmassol> check the error macro
14:24 <vmassol> it's executing some wiki macro
14:24 <sburjan> I know. But something is wrong, because the TestReports were  ok on Fiday, and now they're not. so something changed
14:24 <vmassol> yes sure I could have told you this
14:24 <vmassol> :)
14:24 <vmassol> just look at the page rendered
14:25 <sburjan> how do I do that ?
14:26 <vmassol> open your browser
14:26 <vmassol> and in the URL field , put:
14:26 <vmassol> http://enterprise.xwiki.org/xwiki/bin/view/Main/Roadmap
14:26 <vmassol> you'll see the jira macro is failing
14:27 <sburjan> yes. I saw that. regarding the query param
14:28 <fmancinelli> has joined #xwiki
14:28 <vmassol> it's using a wiki macro
14:28 <vmassol> which I can't locate
14:28 <fmancinelli> Hi all, I am in :)
14:28 <vmassol> ah fmancinelli
14:28 <vmassol> still trying to fix the build
14:29 <fmancinelli> vmassol, you seem happy to see me :)
14:29 <vmassol> just fixed one failure you caused :)
14:29 <fmancinelli> ops. which one?
14:29 <vmassol> (was an easy one to fix)
14:29 <vmassol> the cluster test were not compiling
14:29 <vmassol> http://ci.xwiki.org/view/Functional%20Tests/
14:30 <vmassol> so basically we were quite good but new errors just appeared
14:30 <vmassol> thomas is offlien but he just caused a build failure :(
14:30 <vmassol> (but that's in 4.2)
14:30 <vmassol> then http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-escaping/org.xwiki.enterprise$xwiki-enterprise-test-escaping/1878/testReport/junit/org.xwiki.test.escaping/ObjectAndClassEditorTest/org_xwiki_test_escaping_ObjectAndClassEditorTest/ just appeared
14:31 <vmassol> ok ok
14:31 <vmassol> test-extension still failing
14:31 <vmassol> marius and thomas are on it I think
14:32 <vmassol> and I've been trying to fix the admintest failure on http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-selenium/org.xwiki.enterprise$xwiki-enterprise-test-selenium/1866/testReport/
14:32 <vmassol> the other 2 still need to be fixed
14:32 <vmassol> but they're related and I know why
14:33 <vmassol> fmancinelli: you could start the release
14:33 <vmassol> by applying translations for ex
14:33 <vmassol> fmancinelli: you shoudl start by creating a release plan
14:36 <mflorea> has quit
14:36 <mflorea1> has joined #xwiki
14:36 <mflorea1> has quit
14:37 <fmancinelli> ok, we are releasing 4.3-milestone-1 (I will create a plan here http://dev.xwiki.org/xwiki/bin/view/ReleasePlans/WebHome) please bear with me, this is my first release :)
14:39 <vmassol> sburjan: found it and removed it!
14:39 <sburjan> it was an older jira macro ? even older than mine ?
14:39 <fmancinelli> the id to be put in the textbox for the release plan is 43M1 or 4.3-milestone-1? I think it's the former
14:39 <vmassol> had to write some little script: ;)
14:40 <vmassol> https://gist.github.com/3931320
14:40 <sburjan> ingenious :D
14:41 <vmassol> now hmm I need to restart xwiki.org or register again the component in a script
14:41 <vmassol> I'll restart
14:44 <xwikibot> has joined #xwiki
14:51 <vmassol> fmancinelli: fixed the object, you had put a wrong release manager name
14:52 <fmancinelli> thx
14:53 <fmancinelli> vmassol: let me clone all the projects from xwiki (I usually use a fork from my own account)
14:55 <vmassol> ?
14:56 <vmassol> for applying the translations?
14:57 <fmancinelli> no, for checking other stuff in the release plan (in case I need to fix stuff)
14:57 <vmassol> ok
14:58 <fmancinelli> I think I will need an access on the build machines (reading http://dev.xwiki.org/xwiki/bin/view/ReleasePlans/ReleasePlanHelp)
15:06 <fmancinelli> vmassol, do you use some kind of automatic tool for spotting CLIRR excludes in POMs? Or do you review them one by one trying to find stale excludes?
15:09 <vmassol> find . -name "pom.xml" -exec grep -H "clirr" {} \; or something like thatfm	
15:09 <vmassol> oops
15:09 <vmassol> fmancinelli: see http://dev.xwiki.org/xwiki/bin/view/ReleasePlans/ReleasePlanHelp#HInitialCLIRRcheck
15:09 <vmassol> then you need to review one by one
15:11 <vmassol> jvdrean: bug for you
15:11 <vmassol> when you refactored the panels you broke sometihng
15:11 <vmassol> all the panels are now named "NewPage"
15:11 <vmassol> which make the tests fail
15:12 <Denis> has joined #xwiki
15:13 <jvdrean> vmassol: thanks for the catch, I'll fix it today
15:13 <vmassol> thanks we're waiting for it for the release
15:15 <vmassol> so ASAP would be great :)
15:20 <cdop> is 4.2 usable for production enviroments? ;-) Besides those issues in the releasenote?
15:20 <vmassol> yes
15:21 <cdop> thanks
15:21 <vmassol> but not the ircbot apparently ;)
15:21 <vmassol> fmancinelli: so, once jvdrean fixes the panels issue, we're good to release
15:21 <vmassol> (all tests should be passing)
15:22 <vmassol> with the exception maybe of the extension test
15:22 <vmassol> actually it's passing now
15:22 <vmassol> hmm seems to be flickering
15:23 <vmassol> my guess is that there's osmething asunchronous
15:23 <vmassol> probably the fact that marius made extension retrieval async now
15:23 <vmassol> a pity he's not connected
15:23 <vmassol> yep looks like it: http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-extension/org.xwiki.enterprise$xwiki-enterprise-test-extension/1144/testReport/junit/org.xwiki.test.ui.extension/ExtensionTest/testUpgradeWithMergeConflict/
15:31 <jvdrean> fix committed
15:31 <vmassol> cool that was fast let's see if the tests pass now
15:33 <fmancinelli> I looked at the CLIRR thing... here's what I found:
15:33 <fmancinelli> xwiki-commons/xwiki-commons-core/pom.xml (OK - Remove when the 4.3 final is released)
15:33 <fmancinelli> xwiki-rendering/pom.xml (OK - Remove when 4.3 final is released)
15:33 <fmancinelli> xwiki-platform/xwiki-platform-core/xwiki-platform-rest/xwiki-platformr-rest-server/pom.xml (OK - Moved a lot of packaged to internal, remove on next release)
15:33 <fmancinelli> xwiki-platform/xwiki-platform-core/xwiki-platform-workspace/xwiki-platform-workspace-api (?)
15:33 <mflorea> has joined #xwiki
15:33 <fmancinelli> xwiki-platform/xwiki-platform-core/pom.xml (There's a FIXME)
15:34 <vmassol> fmancinelli:
15:34 <vmassol> xwiki-platform/xwiki-platform-core/xwiki-platform-rest/xwiki-platformr-rest-server/pom.xml
15:34 <vmassol> this is wrong
15:34 <vmassol> and so is:
15:34 <vmassol> xwiki-platform/xwiki-platform-core/xwiki-platform-workspace/xwiki-platform-workspace-api
15:34 <vmassol> all CLIRR excludes must be in a single place as we've agreed
15:34 <vmassol> which is xwiki-platform/xwiki-platform-core/pom.xml for platform
15:35 <fmancinelli> ok fixing it
15:35 <mflorea> vmassol: I updated the extension tests after my latest changes
15:35 <vmassol> fmancinelli: be careful as for workspaces that could be some leftoiver from previously that hasn't been cleaned up
15:35 <vmassol> maybe the previous RM wasn't careful enough
15:36 <vmassol> Enygma`: if you put the clirr excludes there please take note for next time :)
15:36 <vmassol> mflorea: ok but it still looks like some async issue
15:36 <vmassol> missing some wait somewhere probably
15:37 <vmassol> (it got loading when it wanted installed)
15:37 <mflorea> sure, it's flickering, I'm investigting
15:38 <vmassol> ok
15:43 <mrvanes> has joined #xwiki
15:45 <sburjan> vmassol: can I add myself to the Contrib Committers on http://dev.xwiki.org/xwiki/bin/view/Community/HallOfFame ?
15:46 <vmassol> sburjan: definitely, actually you should run the script and update the page :)
15:46 <cdop> has quit
15:47 <sburjan> run the following command on all the Git repositories for the "xwiki-contrib" organization .. wow, this could take a lot of time, a lot of repos :)
15:47 <vmassol> hmm
15:48 <vmassol> s/hmmm//
15:49 <vmassol> actually
15:49 <vmassol> contrib committer has no meaning
15:49 <vmassol> since there's no notion of committer for contrib
15:49 <vmassol> what we need to do is beef up the contributors section
15:50 <vmassol> updating page
15:52 <mflorea> has quit
15:55 <vmassol> sburjan: done
15:55 <vmassol> you can update now
15:56 <fmancinelli> has left #xwiki
15:59 <fmancinelli> has joined #xwiki
16:02 <vmassol> @sorin: my idea was to use this instead;
16:02 <vmassol> http://extensions.xwiki.org/xwiki/bin/view/Extension/GitHub+Application
16:02 <vmassol> (this is why I started it)
16:02 <vmassol> (to replace the current static hall of fame)
16:02 <fmancinelli> vmassol, why you removed yourself? I feel lost now :)
16:03 <vmassol> fmancinelli: becuase you're doing the release :) I can still help you
16:03 <vmassol> so that you get the full credit!
16:04 <vmassol> hmm actually curriki shouldn't be on this page
16:04 <vmassol> since curriki is not following dev.xwiki.org
16:04 <vmassol> removing
16:11 <fmancinelli> what is the profile for not building all the GWT permutations?
16:12 <vmassol> -Pdev
16:12 <vmassol> fmancinelli: http://dev.xwiki.org/xwiki/bin/view/Community/Building#HUsingProfiles
16:13 <fmancinelli> d'oh... I ran mvn with -Ddev that's why it didn't work... nevermind
16:13 <vmassol> :)
16:22 <fmancinelli> for the workspacemanager API CLIRR exclude, I will just move it to the right place and put a comment to remove it on the next final release
16:22 <vmassol> is ti needed for 4.3M1?
16:22 <vmassol> it's not some leftover from previously?
16:22 <vmassol> to check is easy, just rmoeve it and build workspaces
16:22 <vmassol> if it doesn't fail then it's some leftover
16:23 <vmassol> :)
16:23 <vmassol> (fmancinelli)
16:23 <fmancinelli> I did
16:23 <fmancinelli> it fails
16:23 <fmancinelli> so it seems not a leftover but an actual api change (2 methods added)
16:23 <vmassol> ok
16:24 <vmassol> Enygma`: can you please add a comment for the clirr exlcude?
16:24 <vmassol> putting "api changes" is … how to say it… ;) … pretty useless and redundant
16:25 <fmancinelli> I wrote this as a comment: Added methods for creating workspaces and retrieving workspace templates. Remove this check on the next release (i.e., 4.3-M2)
16:25 <vmassol> fmancinelli: this is not what we need to put in teh comment for an exclude
16:25 <vmassol> let me find the doc
16:26 <vmassol> it's not on dev.Xwiki.org, it's in an email...
16:26 <vmassol> so
16:26 <vmassol> what needs to be put is why we accept to break the API
16:26 <vmassol> normally each breakage requires a vote
16:26 <vmassol> and we eight the pros and cons
16:26 <vmassol> *weight
16:27 <vmassol> (http://dev.xwiki.org/xwiki/bin/view/Community/DevelopmentPractices#HBackwardCompatibility)
16:27 <vmassol> so explaining what it does is not the goal
16:27 <vmassol> at least it's not enough
16:27 <fmancinelli> ok, so I have to put there the link to the vote/discussion
16:28 <vmassol> we're on a slippery slope here
16:28 <vmassol> so we need to be extra careful and requiring a good comment is one rule to ensure it's been really carefully discussed
16:28 <vmassol> but we're not very clean since for ex
16:28 <vmassol> the notion of young api is not defined yet
16:29 <vmassol> so anyone can say their api is young and thus explain the breakage with this reason
16:29 <vmassol> in any case it's not for you to explain it but for Enygma`
16:29 <vmassol> since he's theone who added the clirr exclude
16:30 <vmassol> (I guess he's the one who added it :))
16:30 <vmassol> (didn't check)
16:30 <vmassol> requiring a vote email link is a good idea actually
16:32 <vmassol> so IMO we need 2 things: 1) define "young api" and 2) require a vote email link as comment for each clirr exclude
16:32 <vmassol> we have a thread in progress for young apis
16:33 <fmancinelli> Well, we have the same problem for the REST module where I moved a lot of stuff to internal. It was not done as a vote but I got the feedback of Thomas and Jerome...
16:33 <vmassol> that's bad
16:33 <fmancinelli> I am linking that mail in the commen
16:33 <vmassol> I was sure there was a vote
16:33 <vmassol> it's so big
16:33 <vmassol> I'm pretty sure I replied too
16:34 <fmancinelli> hmmm, no
16:34 <fmancinelli> but if you want I can revert the change
16:34 <vmassol> let me check my emails I'm sure I replied
16:34 <vmassol> there were several emails
16:34 <mflorea> has joined #xwiki
16:34 <vmassol> but in any case it should have had a VOTE
16:35 <vmassol> "Well since Fabio says that these classes should have been internal from the beginning and it was a mistake and since Fabio thinks that it's unlikely that anybody would be using them we could make an exception and skip the deprecation and move them to the internal package.
16:35 <vmassol> "
16:36 <vmassol> so no, don't revert now since you've had enough agreement even if "informal"
16:36 <fmancinelli> ok. noted for the next time. sorry
16:36 <vmassol> I can count at least 5 positives agreements and there were no -1
16:38 <vmassol> fwiw http://xwiki.markmail.org/thread/j6qiublcojwj5jws
16:38 <vmassol> I'll resurrect the thread
16:44 <fmancinelli> re workspace exclude, I will just mark the comment as FIXME and let Eduard write a more meaningful comment, so that I can go on
16:44 <vmassol> yep
16:44 <vmassol> let's remember to ping him
16:44 <vmassol> because it could be that he doesn't have a valid reason
16:44 <cdop> has joined #xwiki
16:44 <vmassol> and the api shoudln't have been broken
16:45 <vmassol> so it's important to be sure before the release is over
16:45 <fmancinelli> ok writing a mail
16:45 <fmancinelli> on the other hand... is it really an API breakage? He just added two methods leaving everything else the same
16:47 <vmassol> if clirr says it is, then it is
16:47 <vmassol> is it methods added to an interface?
16:47 <vmassol> seems so
16:48 <vmassol> so it definitely is a breakage
16:48 <vmassol> imagine someone coded an implementation
16:48 <vmassol> and then he upgrades to the new version of xwiki
16:48 <vmassol> his code will fail
16:49 <vmassol> you can imagine an extension on xwiki.org if you rpefer
16:49 <vmassol> and when you try to install it on xwiki 4.3M1 it suddenly deosn't work anymore
16:49 <fmancinelli> oh right, it was an interface indeed
16:49 <vmassol> now
16:50 <vmassol> depending on the definityion of "young api'" workspaces api could be considered as such or not
16:50 <vmassol> ;)
16:50 <vmassol> ATM our best answer is:
16:50 <vmassol> a VOTE with committers deciding whether to break or not
16:50 <Enygma`> vmassol, fmancinelli: the exclude was added by GuillaumeD`s pull request. It allows for having multiple workspace templates. I agree that I should have changed the comment to a more meaningfull one and maybe that there should have been a mail about it, but I assure you that it is a needed feature.
16:51 <vmassol> it's always needed feature that's not the issue
16:51 <Enygma`> but I also assure you that there is no other implementation that is affected by this :)
16:51 <vmassol> the issue is that we globally agree to break an api for a better good
16:51 <vmassol> s/api/people/
16:52 <vmassol>  in this case I think it would be a nobrainer but since it's a serious offense to break an api it's betetr to get an agreement
16:52 <Enygma`> I know this might be good, but isn`t it also slowing progress?
16:52 <vmassol> nobrainer = it's a relatively recent api that not a lof users must be implemeting
16:52 <vmassol> !!!!
16:52 <vmassol> we've  been slowing xwiki progress for 6 years now
16:52 <vmassol> by deciding to evolve xwiki rather than to rerite it
16:52 <vmassol> rewrite
16:53 <vmassol> it was an important decision at the time
16:53 <vmassol> and the consequence is speed of dev ofc
16:53 <vmassol> but the pros is that people can build apps in prod on top of xwiki
16:53 <vmassol> what we all want is JDK8 to be here :)
16:54 <vmassol> since it allows adding default methods for interfaces
16:55 <vmassol> (aka defender methods)
16:56 <vmassol> http://www.javabeat.net/2012/05/virtual-extension-methods-in-java-8/)
16:56 <fmancinelli> Enygma`, I've just pushed the changes... This is the line to fix: https://github.com/xwiki/xwiki-platform/blob/master/xwiki-platform-core/pom.xml#L339
16:57 <Enygma`> so, according to this logic, once a feature is bundled in an XWiki release and it is no longer in the new Roadmap, development of that feature stops completely. Only bugfixes that don`t affect the existing API are accepted and, on a case-by-case manner, some API brakages are accepted for bugfixes/improvements.
16:58 <vmassol> "development of that feature stops completely." not necesarily
16:58 <Enygma`> ah, and new API (new interfaces/services)
16:58 <vmassol> there are lots of ways to add features without breakinfg backward compat
16:59 <Enygma`> but if you had design problems with your initial implementation or you just did not have the time to implement essential features, you will be screwed, and will have to always send mails about it
16:59 <vmassol> hence the notion of "young apis"
16:59 <Enygma`> right
16:59 <vmassol> but yes when it's out of being "young"
16:59 <vmassol> then if you made a design issue you're screwed
16:59 <vmassol> you can check the JDK for example of screw ups
16:59 <vmassol> :)
17:00 <vmassol> the way to get out of screw ups is to do what we've done for rendering
17:00 <Enygma`> what you say fits more with feature-boxing rather than time-boxing
17:00 <vmassol> ie rewrite it in another namespace
17:00 <Enygma`> it`s just not realistic IMO for time-boxing
17:00 <vmassol> and deprecated the old one
17:00 <vmassol> for me that's completely unrelated
17:01 <vmassol> can't even think of a common point
17:01 <Enygma`> what I meant was that when you have a time limit, your design will also need to be limited to be able to implement it.
17:02 <Enygma`> but anyway, I agree that refactoring at some point is the answer
17:02 <vmassol> first you need to realize we're talking about apis here not features
17:02 <vmassol> for the end user it has no impact
17:03 <vmassol> so the good practice is to make somehow the api as young (internal package ATM is all we really have for that)
17:03 <fmancinelli> I suppose that I should move to 4.3-M2 all the open JIRAs that are marked as fix 4.3-M1
17:03 <vmassol> fmancinelli: yes but please check that there are been no commits
17:03 <Enygma`> right, so you will end up with an API that is as expressive as possible, but not all methods will be implemented because you did not have the time for them. But you will implement them eventually
17:03 <vmassol> (commits tab)
17:03 <fmancinelli> ok
17:04 <vmassol> sometimes devs forget to close issues
17:04 <vmassol> Enygma`: YAGNI principle
17:04 <vmassol> aka overdesign
17:05 <vmassol> and the answer is indeed refactoring
17:05 <Enygma`> agreed, but then you have backwards compatibility
17:05 <Enygma`> :)
17:05 <Enygma`> right
17:05 <vmassol> no
17:05 <vmassol> only if you put them publicly available for users to write against
17:05 <vmassol> hence the reason we have internal packages for ex
17:05 <vmassol> this means: don't touch this, it can change anytime, we don't gurantee backward compat
17:06 <Enygma`> yes, so indeed the internal APIs solution does seem good... even if you might have some devs not able to use your APIs untill you decide to "publish" them
17:06 <vmassol> it's better to code defensively
17:06 <vmassol> which means close api and then open when user ask for it
17:06 <Enygma`> so people really needing them will use them and then refactor when they become public... but you will end up with a fake solution, since people will *need to* refactor anyway :)
17:06 <Enygma`> but at least you can say that it`s their fault :)))
17:06 <vmassol> yep
17:07 <vmassol> you can tell them "I told ya!" ;)
17:07 <Enygma`> I call that cheating, but anyway :)
17:07 <vmassol> no
17:07 <vmassol> because it's visible
17:07 <vmassol> it's a contract
17:07 <vmassol> of course
17:07 <vmassol> if 100% of your software is internal
17:07 <vmassol> then it just mean you're not a dev platform
17:07 <vmassol> that's all
17:07 <vmassol> no cheating
17:07 <Enygma`> I would have settled with a warning on the public API that it`s young, instead of the whole internal sharade
17:08 <vmassol> Enygma`: the thred has not been closed
17:08 <vmassol> we are currently undecided
17:08 <vmassol> http://markmail.org/thread/hzhoncujzn37k7qh
17:08 <vmassol> the only rule we have now
17:08 <vmassol> is that any CLIRR exclude must require a VOTE
17:09 <vmassol> this has been voted and is documented
17:09 <vmassol> http://dev.xwiki.org/xwiki/bin/view/Community/DevelopmentPractices#HBackwardCompatibility
17:09 <Enygma`> yes, I know
17:09 <vmassol> "We never remove APIs from the legacy modules by default. However if we really need to do so (for some technical reason for example), we do it on a case by case basis with a VOTE."
17:09 <vmassol> so at least we control our breakages ATM
17:09 <vmassol> even though we still miss a clear guideline for young apis
17:10 <vmassol> we control = we're supposed to control
17:10 <vmassol> because there are still clirr excluides without votes
17:10 <Slashman_> has quit
17:10 <Enygma`> re the thread, what about an option 4) using the @Deprecated annotation in a semantically-flawed manner? :)
17:10 <vmassol> feel free to reply
17:10 <Enygma`> k
17:11 <Slashman> has quit
17:11 <fmancinelli> There are 6 issues marked 4.3-M1 with commits...
17:13 <vmassol> oops that's a lot of open issues
17:14 <vmassol> we need to ping devs about them
17:14 <vmassol> can you list them?
17:14 <fmancinelli> http://jira.xwiki.org/secure/IssueNavigator.jspa?reset=true&jqlQuery=Project+in+%28%22XWiki+Commons%22%2C+%22XWiki+Rendering%22%2C+%22XWiki+Platform%22%2C+%22XWiki+Enterprise%22%2C+%22XWiki+Enterprise+Manager%22%29+and+fixVersion+in+%28%224.3-milestone-1%22%29+and+resolution+in+%28%22Unresolved%22%29
17:14 <fmancinelli> I moved the others (without commits) to 4.3-M2
17:14 <vmassol> ehcking mine
17:15 <vmassol> mine is ok
17:15 <vmassol> It's been rolledback
17:15 <vmassol> fixed fix for
17:16 <vmassol> marius has comlitted in a branch fmancinelli
17:16 <vmassol> moving to 4.3M2
17:16 <vmassol> smae for rthomas
17:16 <vmassol> fixing
17:17 <vmassol> same for edy
17:18 <vmassol> only 1 left
17:18 <fmancinelli> so there's thomas left: http://jira.xwiki.org/browse/XCOMMONS-204?page=com.atlassian.jira.plugins.jira-bitbucket-connector-plugin:bitbucket-commits-tabpanel
17:18 <vmassol> we need to ask thhomas
17:18 <vmassol> nope
17:18 <vmassol> 204 is ok
17:18 <vmassol> http://jira.xwiki.org/browse/XCOMMONS-280 is the one left
17:19 <vmassol> 204 was rolledback
17:19 <vmassol> commented in it
17:19 <vmassol> (for 280)
17:21 <fmancinelli> I see.
17:26 <jvdrean> fmancinelli: I'd like to commit an tiny minor improvement for http://jira.xwiki.org/browse/XWIKI-7681?focusedCommentId=72428#comment-72428 is it too late ?
17:26 <jvdrean> it's a CSS selector modification that fixes the issue raised by Caty, vmassol WDYT ?
17:27 <vmassol> jvdrean: +1 but provided you follow the functional tests on the CI :)
17:27 <vmassol> speaking of which, let me check the results of your ealier fix
17:28 <vmassol> yeah http://ci.xwiki.org/job/xwiki-manager-test-selenium/ :)
17:28 <vmassol> ah no
17:28 <vmassol> wrong project
17:28 <vmassol> :(
17:28 <fmancinelli> jvdrean, ok
17:28 <vmassol> still building: http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-selenium/1870/
17:29 <vmassol> ouch
17:29 <vmassol> http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-ui/
17:30 <vmassol> "java.net.ConnectException: Connection refused
17:30 <vmassol> "
17:30 <vmassol> ah no that's not the reason
17:31 <vmassol> hmm strange
17:31 <vmassol> so it's failing on:        Assert.assertTrue(getUtil().isInCreateMode());
17:32 <sburjan> has quit
17:33 <jvdrean> pushed, since the modification is only in the CSS, not the HTML, it shouldn't be a problem with the func tests
17:34 <vmassol> fmancinelli: you should use —rebase when pulling ;)
17:35 <vmassol> (I see a merge in your commit)
17:36 <vmassol> jvdrean: http://ci.xwiki.org/view/Functional%20Tests/job/xwiki-enterprise-test-selenium/ :)
17:36 <vmassol> you should look at functional tests more often jvdrean, this one has been failing for a week or so
17:36 <vmassol> :)
17:37 <vmassol> it's a pain for the RM
17:37 <vmassol> to have to go over all the mistakes done by committer
17:37 <vmassol> committers
17:37 <jvdrean> yep, I quite remember how it feels
17:37 <vmassol> jvdrean: you should aso have received daily emails about this one
17:37 <vmassol> telling you that you broke the build no?
17:37 <vmassol> hmm maybe not for functionla tests
17:38 <jvdrean> I didn't see emails about that but I might have a filter, let me see
17:38 <vmassol> I hope you don't have any filter
17:38 <vmassol> ie mails sent to you directly and not the list must not be filteree
17:38 <vmassol> *filtered
17:39 <jvdrean> label:notifications to:[email protected] returns nothing
17:39 <fmancinelli> vmassol, ah. didn't know it. I use rebase before pushing to clean things up. didn't know it worked also when pulling.
17:40 <fmancinelli> vmassol, thomas just arrived at the office :)
17:40 <vmassol> it rewrite your local changes on top of the already pushed changes
17:40 <vmassol> oterhwise it creates a branch and merge it later on
17:40 <vmassol> it's our best practice on this project
17:40 <vmassol> to always use --rebase
17:40 <vmassol> to remove those spurious merges
17:41 <vmassol> which are unnecessary and pollute our history tree
17:41 <jvdrean> from:[email protected] to:[email protected], no results
17:42 <vmassol> jvdrean: let me find some because I'm sure I was some recently
17:43 <vmassol> you were in it, lucky you
17:43 <vmassol> ;)
17:43 <vmassol> s/ in it/not in it
17:43 <vmassol> I think I set this up only for commons, rendering and platform so far
17:43 <vmassol> but that doesn't excuse you from checking ci.xwiki.org after your commits
17:44 <vmassol> ah found one
17:44 <vmassol> 7th of September
17:44 <vmassol> "You broke the build! Fix it now!"
17:44 <vmassol> can you see it jvdrean?
17:45 <jvdrean> nope
17:45 <vmassol> jv at drean dot org
17:46 <tmortagne> has joined #xwiki
17:49 <jvdrean> indeed, tagged as spam
17:49 <jvdrean> fixing it
17:57 <evalica> has quit
18:01 <Enygma`> has quit
18:08 <fmancinelli> There are missing documentation links in issues: http://jira.xwiki.org/secure/IssueNavigator!executeAdvanced.jspa?jqlQuery=Project+in+%28%22XWiki+Commons%22%2C+%22XWiki+Rendering%22%2C+%22XWiki+Platform%22%2C+%22XWiki+Enterprise%22%2C+%22XWiki+Enterprise+Manager%22%29+and+fixVersion+in+%28%224.3-milestone-1%22%29+and+resolution+in+%28%22Fixed%22%29+and+%28%28Documentation+is+EMPTY%29+or+%28%22Documentation+in+Release+Notes%22+
18:09 <vmassol> 11 that's a lot
18:09 <vmassol> AndreasJonsson: you have 2
18:09 <fmancinelli> fixed mine :)
18:09 <vmassol> tmortagne: 4
18:10 <vmassol> jvdrean: 3
18:10 <vmassol> cjd: 1
18:10 <tmortagne> vmassol: I'm on it
18:14 <tmortagne> vmassol: I had more than 4 actually but you checked too late ;)
18:15 <vmassol> hehe
18:15 <vmassol> nothing to be proud of! :)
18:15 <sburjan`> has joined #xwiki
18:21 <vmassol> fmancinelli: I think http://jira.xwiki.org/browse/XWIKI-7966 is missing a link to the documentation too
18:21 <vmassol> also fmancinelli you should not put absolute URL on xwiki.org: http://jira.xwiki.org/browse/XWIKI-7966
18:21 <vmassol> oops wrong paste
18:21 <vmassol> I meant: http://www.xwiki.org/xwiki/bin/view/ReleaseNotes/ReleaseNotesXWiki43M1#HRESTAPIimprovements
18:22 <vmassol> fmancinelli: the reason is that these won't be refactored automatically when a rename happens
18:22 <vmassol> in addition it's not really nice looking ;)
18:23 <vmassol> (you can create links on labels in wiki ;))
18:23 <vmassol> but in any case the most important is that the ref doc is missing
18:37 <fmancinelli> vmassol, ok fixed. for the ref doc I will document it in the REST API page.
18:40 <vmassol> fmancinelli: thanks, yep
19:08 <fmancinelli> The current platform build is failing on a lot of GWT tests... Thomas says that it's caused by the agent build agent
19:08 <fmancinelli> What to do? Start another build on a different agent to see if everything is actually ok?
19:10 <tmortagne> I disabled KVM-Agent-CentOS-GUI 
19:10 <tmortagne> I don't even know what this agent is supposed to be...
19:13 <vmassol> wow this looks cool: http://labs.carrotsearch.com/junit-benchmarks.html
19:16 <coud5> has joined #xwiki
19:23 <mflorea> has quit
19:25 <abusenius> has joined #xwiki
19:29 <vmassol1> has joined #xwiki
19:29 <vmassol> has quit
19:36 <vmassol1> has quit
19:47 <fmancinelli> going back home. Thomas fixed a problem on the build machine and restarted the build. this time we should have a complete build hopefully without errors
19:47 <fmancinelli> I've pinged the people who need to fix the documentation. Please do it ASAP
19:59 <tmortagne> has quit
20:07 <vmassol> has joined #xwiki
20:07 <fmancinelli> has quit
20:16 <vmassol> has quit
20:32 <mflorea> has joined #xwiki
20:33 <coud6> has joined #xwiki
20:35 <coud5> has quit
20:40 <vmassol> has joined #xwiki
20:58 <mflorea> has quit
20:59 <mflorea> has joined #xwiki
21:05 <vmassol> has quit
21:08 <vmassol> has joined #xwiki
21:11 <coud7> has joined #xwiki
21:13 <coud6> has quit
21:50 <mflorea> has quit
21:53 <mrvanes> has quit
21:55 <fmancinelli> has joined #xwiki
22:11 <fmancinelli_> has joined #xwiki
22:14 <vmassol> thanks AndreasJonsson :)
22:22 <fmancinelli> the build looks fine (platform)
22:31 <cdop> ok, xwiki (4.2) completely fresh installed on an ubuntu tomcat server. worked pretty much out of the box.
22:31 <cdop> but now: home&bed :)
22:31 <cdop> has quit
22:33 <vmassol> has quit
22:38 <coud7> has quit
23:05 <vmassol> has joined #xwiki
23:06 <sburjan`> vmassol: I was thinking (but unable to test ATM). Maybe the "Remember username" feature is also broken as the recover password was
23:08 <sburjan`> no, actually that does not send an e-mail
23:11 <vmassol> has quit
23:36 <vmassol> has joined #xwiki
23:40 <vmassol> has quit
23:47 <abusenius> has quit
23:51 <SvenDowideit> has quit